|Publication number||US7277509 B2|
|Application number||US 10/418,302|
|Publication date||Oct 2, 2007|
|Filing date||Apr 18, 2003|
|Priority date||Feb 10, 2003|
|Also published as||US20040156422|
|Publication number||10418302, 418302, US 7277509 B2, US 7277509B2, US-B2-7277509, US7277509 B2, US7277509B2|
|Original Assignee||Nokia Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (9), Referenced by (8), Classifications (14), Legal Events (6)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application claims priority of U.S. Provisional Patent Application Ser. No. 60/445,814, entitled “Low-Complexity Frequency-Offset Correction Method,” filed on Feb. 10, 2003, the contents of which are hereby incorporated by reference.
1. Field of the Invention
The present invention relates to the correction of carrier frequency offsets in a mobile communication network.
2. Description of the Related Art
The signal from the multiplier 10 is also supplied to a further integrator 12, the integrated control symbols being supplied to a channel estimation block 14. The channel estimation block 14 estimates the complex channel coefficient of the radio channel using pilot signal information (or a training sequence in TDMA receiver), and provides a channel estimate for removing the effects of the radio channel from the received signal. The signal from the integrator 18 is multiplied by the complex conjugate of the channel estimate at multiplier 20. The resulting output includes recovered data in the form of soft symbol data with reliability information associated with the recovered data.
The output is further transformed into a real signal in block 22, and then supplied to subsequent de-coding operations. It will be appreciated that the imaginary part may also include data, in which case the invention could likewise be applied to the imaginary part. A problem that exists with the system illustrated in
Another cause for rotation of the signal is the so-called Doppler effect. This means that the length of the radio signal path between the mobile station and the base station changes when the mobile station moves. This causes a Doppler shift in the spectrum in the received signal. A mobile station synchronises its system clock according to the received signal from the base station. For a moving mobile, the Doppler effect will modify the observed carrier frequency. Therefore the mobile will end up transmitting the wrong carrier frequency. When the base station receives the signal the Doppler effect has again modified the carrier frequency in the same direction so the base station observes a carrier frequency offset that is two times the Doppler shift.
In mobile station receivers, the frequency offset may be detected and the frequency of the local oscillator can be controlled to remove the offset. However, this relies on information being available as to the exact amount of the frequency offset. Another problem is that the resolution of the oscillator frequency adjustment is usually too coarse to adequately compensate the frequency offset. In base station receivers it is not possible to adjust the local oscillator frequency, because the local oscillators are common for several channels and the frequency offset is usually different for signals that were received from different mobile stations.
A significant problem with the carrier frequency offset is that it shifts the power spectrum of the received signal. Because of the Doppler shift, the spectrum is no longer symmetric so, a channel estimator with real valued filter coefficients no longer works in an optimum way.
One existing approach to deal with this problem is that the carrier frequency offset is not corrected at all. The main idea here is to choose filter coefficients with a large enough passband so that the actual signal is not filtered away in channel estimation. This has very poor performance with large carrier frequency errors.
In the system illustrated in
Another possibility for correcting the carrier frequency offset is to monitor the phase of the channel estimate and generate a complex phasor on the basis of successive phase values. The resulting phasor is then used to correct the received base band signal before or after channel estimation. The carrier frequency offset is removed by de-rotating samples both after descrambling with the long code and descrambling with the short code. A disadvantage with this method is that the computations which are required depend on the data rate. Also, the number of computations which is required is so high that it is not feasible to implement this with DSP software in a practical system. Such a system is discussed in EP-A-1160981.
It is an object of the invention to correct frequency offsets in radio carriers, while reducing the computational complexity required to do so.
According to one aspect of the invention there is provided a receiver for use in a mobile communications network, the receiver having an input for receiving radio signals including control symbols and data symbols; a channel estimator arranged to use the control symbols to provide a channel estimate for correcting received data symbols; a first offset corrector for de-rotating the control symbols prior to their use in providing the channel estimate; a second offset corrector for rotating the channel estimate prior to its use in correcting the received data symbols; and an offset estimator arranged to generate estimates of frequency offset in the received radio signal, for use in the first and second offset correctors. The offset can be estimated based on the control symbols.
According to another aspect of the invention there is provided a method for correcting frequency offset in a received radio signal, the method having the steps of using control symbols in the received radio signal to provide a channel estimate; generating an estimate of frequency offset in the received radio signal; de-rotating control symbols prior to their use in providing the channel estimate; rotating the channel estimate; and using the rotated channel estimate to correct data symbols in the received radio signal; wherein the steps of de-rotating the control symbols and rotating the channel estimate are carried out using estimates of frequency offset.
A radio signal used in a mobile communication network can have a sequence of time slots, each time slot including data symbols and control symbols. The receiver can include a filter, which is arranged to generate a first set of control samples representing the control symbols for each time slot. It will be appreciated that data symbols and control symbols can be sent on the same channel, or on separate channels.
In one embodiment, the offset estimator is arranged to produce one estimate of frequency offset for each time slot based on the first set of control samples. This one estimate is used to update a phasor for both the first and second offset correctors.
In another embodiment, a second filter can be arranged to receive the first set of control samples and to produce a second set having a different number of control samples for supply to the first offset corrector. A number of frequency estimates is produced for that time slot to correct the number of control samples in the second set supplied to the first offset corrector. It is also possible to include a first interpolator for receiving values representing the channel estimate from the channel estimator and arranged to generate interpolated values between the received values, the received values and the interpolated values being supplied to the second offset corrector. In that case, a second set of frequency offset estimates can be produced for that time slot matching in number the number of received values and interpolated values to be rotated at the second offset corrector.
If necessary, a third filter can be connected to receive de-rotated control samples from the first offset corrector and to supply samples to the channel estimator at a sampling rate different from that of the de-rotated control samples.
Likewise, a second interpolator can be provided for receiving rotated channel estimate values from the second offset corrector and for providing interpolated values from said values and for supplying said received values and said interpolated values for correcting received data symbols.
The receiver can implement a feed back or a feed forward offset correction system.
In a feed forward offset correction system, the first set of control samples is de-rotated at the first offset corrector using the one estimate produced for the same time slot as for the control samples.
In a feed back offset correction system, the offset estimator is arranged to generate said one estimate from control symbols of a first time slot and the first offset corrector is arranged to de-rotate the first set of control samples generated for a second time slot based on said one estimate produced for the first time slot.
As described more particularly with reference to the following exemplary embodiment, this invention corrects the carrier frequency offset from a received radio signal. This is done in a preferred embodiment by estimating the carrier frequency offset from the complex auto-correlation of the control symbols. Using the carrier frequency offset estimate, the carrier frequency offset is removed from the control channel by de-rotating the control symbols. After the control symbols have been de-rotated, normal channel estimation can be carried out using the de-rotated control symbols. Before the channel estimates are used to correct data symbols, however, the channel estimates need to be rotated to the correct phase.
Carrier frequency offset in a received signal causes a static rotation of the received signal. By taking the static rotation into account in the channel estimation, it is possible to totally remove the carrier frequency offset. In this way, the carrier frequency offset in the received signal does not degrade the performance of the receiver.
In any situation where the carrier frequency offset is corrected, the performance of the receiver is approximately the same, or rather it depends on how often updates are made. It is important to try and reduce as far as possible the required computational complexity needed in order to remove the carrier frequency offset, in particular because this controls how often estimates can be made and this therefore has an impact on performance.
The technique described herein has the following benefits over earlier known techniques.
It is possible to implement the technique with a fixed amount of computations for each round, because it does not depend on the data rate because only the control symbols are used to generate the frequency offset estimate and only the control signals are de-rotated using that estimate (prior to channel estimation).
The channel estimation can be carried out with real valued filter coefficients. This means only roughly half of the computations, which are required in a situation where channel estimation requires the use of complex valued filter coefficients.
The technique described in the following allows for control symbols to be de-rotated directly, or for control symbols at a reduced sampling rate to be de-rotated. Therefore the sample rate can be varied.
The technique described in the following is such of that the computational complexities are at a level where it is feasible to implement the method on DSP software. The method could also be implemented on ASIC.
For a better understanding of the present invention and to show how the claim may be carried into effect reference will now be made by way of example to the accompanying drawings:
The system which has just been described is an example of the so-called feed forward system. An alternative feed back system will now be described, still with reference to
The system is initialized by setting the carrier frequency offset estimate Fest to zero and the correction to the frequency offset estimate (Fest) also to zero. During the first round, the control symbols are rotated using an arbitrary phasor at the multiplier 40′ which takes the place of the multiplier 40. The rotated values are supplied to the prefilter 36. The output from the prefilter 36 is used to estimate the correction term Fest that is used to update the frequency offset estimate Fest for the next round. The frequency offset estimate Fest is used to produce phasors for the different control symbols. Assuming that the prefilter 36 changes the sampling rate to one sample per slot, the carrier frequency offset estimate Fest gives an indication of how much the phasor should change for the next time slot. The channel estimation block 42 uses the values from the prefilter 36. Note that the sampling rate of the derotation and rotation is different here in the feedback system so that different phases could be used for different control symbols.
In this case, the phasor which is used to rotate the control symbols prior to the filter is based on the preceding control symbols which were supplied in a preceding time slot. In other respects the system is the same as that described with reference to
On the output side of the channel estimator 2 a first interpolation block 76 is provided, the purpose of which is to provide interpolated values based on the channel estimates provided by the channel estimator 42 to increase the number of channel estimates per slot. The actual and interpolated channel estimate values are supplied to the multiplier 46. The carrier frequency offset estimation and phase rotation block 38 produces a number of phasors 5 per slot which matches the number of interpolated values per slot generated by the first interpolation block 76. These phasors are supplied via the delay 48 to the multiplier 46 where they are multiplied by the respective actual and interpolated values generated by the interpolation block 76. The thus rotated channel estimate values are supplied to a second interpolation block 78 which generates further interpolated values to the desired output sampling rate.
At step S14 the despread data symbols are multiplied with the final channel estimates and supplied to a combiner which combines (step S15) the results of this rake finger with other rake fingers. The combined output is converted to a real signal at step S16 and output at step S17.
One having ordinary skill in the art will readily understand that the invention as discussed above may be practiced with steps in a different order, and/or with hardware elements in configurations which are different than those which are disclosed. Therefore, although the invention has been described based upon these preferred embodiments, it would be apparent to those of skill in the art that certain modifications, variations, and alternative constructions would be apparent, while remaining within the spirit and scope of the invention. In order to determine the metes and bounds of the invention, therefore, reference should be made to the appended claims.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US5799010 *||Jun 27, 1996||Aug 25, 1998||Interdigital Technology Corporation||Code division multiple access (CDMA) communication system|
|US6229843 *||Nov 22, 1999||May 8, 2001||Interdigital Technology Corporation||Pilot adaptive vector correlator|
|US6363102 *||Apr 23, 1999||Mar 26, 2002||Qualcomm Incorporated||Method and apparatus for frequency offset correction|
|US6456608 *||Jan 18, 2001||Sep 24, 2002||Interdigital Technology Corporation||Adaptive vector correlator using weighting signals for spread-spectrum communications|
|US6697350 *||Jan 18, 2001||Feb 24, 2004||Interdigital Technology Corporation||Adaptive vector correlator for spread-spectrum communications|
|US6983009 *||Jan 18, 2001||Jan 3, 2006||Interdigital Technology Corporation||Median weighted tracking for spread-spectrum communications|
|US20030076812 *||Feb 22, 2001||Apr 24, 2003||Benedittis Rosella De||Method for optimizing the random access procedures in the cdma cellular networks|
|US20030193914 *||Jan 10, 2001||Oct 16, 2003||Gary Lomp||Rapid acquisition spreading codes for spread-spectrum communications|
|EP1160981A2||May 22, 2001||Dec 5, 2001||Nokia Mobile Phones Ltd.||Method and arrangement for reducing frequency offset in a radio receiver|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7434145 *||Apr 1, 2004||Oct 7, 2008||Qualcomm Incorporated||Extracting soft information in a block-coherent communication system|
|US8196000||Jun 11, 2007||Jun 5, 2012||Qualcomm Incorporated||Methods and apparatus for interleaving in a block-coherent communication system|
|US8619841 *||Nov 13, 2012||Dec 31, 2013||Marvell International Ltd.||Transceiver with carrier frequency offset based parameter adjustment|
|US9215712||Feb 4, 2013||Dec 15, 2015||Broadcom Corporation||Method and system for iterative multiple frequency hypothesis testing with Cell-ID detection in an E-UTRA/LTE UE receiver|
|US20040157626 *||Feb 9, 2004||Aug 12, 2004||Vincent Park||Paging methods and apparatus|
|US20040196927 *||Apr 1, 2004||Oct 7, 2004||Hui Jin||Extracting soft information in a block-coherent communication system|
|US20070060175 *||Nov 15, 2006||Mar 15, 2007||Vincent Park||Paging methods and apparatus|
|US20070234175 *||Jun 11, 2007||Oct 4, 2007||Qualcomm Incorporated||Methods and apparatus for interleaving in a block-coherent communication system|
|U.S. Classification||375/344, 375/150, 375/149, 375/147, 375/148, 375/E01.002|
|International Classification||H04L27/06, H04L27/00, H04B1/707|
|Cooperative Classification||H04L2027/0091, H04B1/707, H04L2027/0065, H04L2027/0046|
|Jul 31, 2003||AS||Assignment|
Owner name: NOKIA CORPORATION, FINLAND
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:LILJESTROM, HENRIK;REEL/FRAME:014343/0065
Effective date: 20030710
|Feb 21, 2008||AS||Assignment|
Owner name: NOKIA SIEMENS NETWORKS OY, FINLAND
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:NOKIA CORPORATION;REEL/FRAME:020550/0001
Effective date: 20070913
Owner name: NOKIA SIEMENS NETWORKS OY,FINLAND
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:NOKIA CORPORATION;REEL/FRAME:020550/0001
Effective date: 20070913
|Mar 28, 2011||FPAY||Fee payment|
Year of fee payment: 4
|May 15, 2015||REMI||Maintenance fee reminder mailed|
|Oct 2, 2015||LAPS||Lapse for failure to pay maintenance fees|
|Nov 24, 2015||FP||Expired due to failure to pay maintenance fee|
Effective date: 20151002